Hi Loes,
One thing that always holds is that you never do further manipulations
with statistical images (like T and F). The only thing that you can do
with those is threshold them to find significant blobs. So what you
could take to the second level is contrast images, the differences
between activation and baseline. I don't think there is a reason to do
this though as you will just make your analysis much less powerful
since the baselines introduce more noise into the images. There is no
problem to compare your activation windows directly.
Best,
Vladimir
On Thu, Jul 5, 2012 at 1:37 PM, Loes Koelewijn <[log in to unmask]> wrote:
> Hi Vladimir,
>
> I have another question. I've run an inversion (without TF contrast) on
> unaveraged data, subsequently focussing on induced responses within a
> frequency band. I've created images of the outcome for all subjects &
> conditions for a time window, and now wish to turn to second level analysis.
> I was wondering whether this activity should not be contrasted to activity
> during a baseline period, i.e., create the same images for a baseline time
> window, then contrast each corresponding post-stimulus window in a t test,
> and then perform second level analysis on the t images. Or am I making this
> unnecessarily complex and could one just turn to direct second-level
> analysis of the active time windows over conditions?
>
> Many thanks for some help,
>
> Loes
>
> --
> Loes Koelewijn
> PhD Candidate
> Macquarie Centre for Cognitive Science (MACCS)
> ARC Centre of Excellence in Cognition and its Disorders (CCD)
> Macquarie University
> Sydney NSW 2109
> Australia
>
> Ph: +61 2 9850 4135
> Fax: +61 2 9850 6059
> email: [log in to unmask]
>
|